diff --git a/paper.md b/paper.md index 5ab9c27c79dfd27562debb91abd451fb70b1485a..0d7b040dd2b7be06ce343728bd9bfa169b584af6 100644 --- a/paper.md +++ b/paper.md @@ -54,19 +54,17 @@ To add new neighbourhoods, it is necessary to store them as GeoJSON and make the The current version of Predihood is bundled with data from France using the [mongiris](https://gitlab.liris.cnrs.fr/fduchate/mongiris) project. It includes about 50,000 neighbourhoods with 640 indicators, and 300 neighbouhoods were annotated by social science researchers (one to two hours per neighbourhood to investigate building and streets pictures, parked cars, facilities and greens areas from services such as Google Street View). -## Predicting neighbourhoods +## Predicting environment -There are mainly four steps: producing supervised neighbourhoods, collecting data about neighbourhoods, compute datasets and finally running algorithms to predict environment. - - - -The third step aims at computing datasets that will aggregate aforementioned data. A dataset looks like Figure 1 and is composed of the code INSEE of the neighbourhood (grey column), its indicators (yellow columns) that have been normalized by density of population (green column) and the assessment of social science researchers for the six environment variables (blue columns). As a reminder, our approach Predihood aims at automatically filling question marks for neighbourhoods that are not yet assessed. +Machine learning algorithms need a dataset, as illustrated by Figure 1. In Predihood, a dataset is composed of the identifier of the neighbourhood (grey column, named code INSEE), its indicators (yellow columns) that have been normalized by density of population (green column) and optionnaly the assessment of social science researchers for the six environment variables (blue columns). The objective of Predihood is to fill automatically question marks for neighbourhoods that are not yet assessed.  -The last step predicts the environment of any neighbourhood in France. Because neighbourhoods are represented by hundreds of indicators, a selection process selects subsets of relevant indicators. These subsets are called _lists_ and contain from 10 to 100 indicators. They are used in the Predihood interface to predict environment. +To perform prediction, a selection process first selects subsets of relevant indicators. These subsets, called _lists_, contain from 10 to 100 indicators. Predihood provides a cartographic interface based on [Leaflet](https://leafletjs.com/) and [Open Street Map](https://www.openstreetmap.org/), as shown in Figure 2. It enables to search for a neighbourhood and predict its environment by selecting an algorithm. The current version of Predihood currently includes 8 predictive algorithms from [scikit-learn](https://scikit-learn.org/) (e.g., Random Forest). + + -## Algorithmic interface +## Adding new algorithms Because the prediction of these variables is a complex task, we have to test several algorithms to compare results. In order to facilitate the tuning and the using of the algorithms, Predihood proposes a generic and easy-to-use interface for algorithms. This interface is based on [Scikit-learn](https://scikit-learn.org/stable/) algorithms but can handle hand-made ones. To implement your own algorithm and test it on our dataset, follow these steps: @@ -109,16 +107,10 @@ class MyOwnClassifier(Classifier): After that, your algorithm is ready to be used in Predihood. -Figure 2 shows the generic interface of Predihood for tuning algorithms. The left panel allows to tune parameters and hyper parameters, such as training and test sizes. On the right, the table illustrates the accuracies obtained for each list (generated during the selection process) and each environment variable. You can export these results by clicking on the download icon. +Figure 3 shows the generic interface of Predihood for tuning algorithms. The left panel allows to tune parameters and hyper parameters, such as training and test sizes. On the right, the table illustrates the accuracies obtained for each list (generated during the selection process) and each environment variable. You can export these results by clicking on the download icon.  -## Cartographic interface - -Figure 3 exposes the cartographic interface of Predihood, used mostly by people who search for a new living place. By searching an area in the inputs on the left and then clicking on neighbourhoods, you will be able to choose an algorithm to predict environment variables of the chosen neighbourhood. For beginners, `Random Forest` classifier is recommended. For example, Alice is an IT commercial and has been recruited for a mission in Lyon for 6 months before going back to Paris. She compares easily many neighbourdhoods in the CBD (Central Business District) of Lyon and chooses the "Part-Dieu" neighbourhood. - - - # Mentions of Predihood Our approach Predihood has been presented during the DATA conference [@barretpredicting].